Functional brain imaging with positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) has provided unique new insights into the functioning of the human brain. The brain signals detected by these imaging devices result from a combination of changes in local brain circulation and energy metabolism that offer a unique opportunity, heretofore unexplored, to examine synaptic function in the context of the synaptic homeostasis hypothesis proposed in this application. Two features of the imaging signals are particularly important in this regard. First, imaging based on circulatory and metabolic changes associated with brain function is singularly sensitive to changes in synaptic activity. This reflects the fact that dendrites and axon terminals have high surface-to-volume ratios making synaptic activity metabolically very demanding. Second, glutamatergic neurotransmission appears to account for a very large fraction of this metabolic activity and is uniquely identified in imaging signals due to the use of aerobic glycolysis by astrocytes to remove it from synapses. Because glutamate has been specifically identified as having an important role in learning and memory, this interesting combination of factors places brain imaging with PET and fMRI in a unique position to test important aspects of the synaptic homeostasis hypothesis. In the proposed experiments we will utilize both PET and fMRI along with EEC. We hypothesize that learning will be associated with persistent, regionally specific increases in brain aerobic glycolysis in the resting state (awake, lying quietly with eyes closed) which will be manifest not only as an increase in glucose metabolism that is greater than any increase in oxygen consumption as measured with PET but also in an increase in the spontaneous fluctuations in the fMRI BOLD signal, an important indicator of the intrinsic activity and organization of the brain. Further, we predict that these learning induced changes.will return to baseline following a night of normal sleep but will not do so if SWS is selectively disrupted. We believe that these experiments will provide critical new information relevant to our understanding of the synaptic homeostasis hypothesis. It is important to note that this research is proposed to take place in the context of a new and important collaboration among investigators with highly complementary talents and interests. This is a unique opportunity for all concerned.